Go Big (Data) or Go Home

News flash: those who invest more than others tend to yield bigger gains.
And not just in dollar terms, but in organizational commitment too — the “adapt or die” mentality.
Summary article by by Pam Baker in FierceBigData and original article by Satya Ramaswamy in Harvard Business Review blogs.
Emphasis in red added by me.
Brian Wood, VP Marketing

The top 8 traits of big data winners

The learning curve in big data is either long and winding or sharp and dangerous, depending on the path you take. It’s no secret that many companies drive right off the cliff either way. But those that make that curve race to the finish line and the winner’s circle.
There doesn’t appear to be much of a middle ground at this point. Most companies either make that curve or they don’t. Others have yet to even get on the road.
Tata Consultancy Services surveyed 1,217 executives from companies with more than $1 billion in revenues in a dozen global industries to see if they could determine why some companies make it to the winners circle and others never finish the race.
They found four common traits in big data winners: The leaders are more Internet-centric; they mined data for multiple reasons and uses; they are more aggressive in exploiting unstructured and external data; and, they centralize their big data analysts rather than embed them in business functions.
From my own experience in observing the field, I would add four additional traits found in big data winners: their analysts are uncommonly knowledgeable about the company; analysts are partnered with highly accomplished business users who are also uncommonly familiar with the company’s business and industry; their goal is never to reduce costs but to improve profits and growth; and, they know exactly what they seek to accomplish with big data from the start.
In other words, the big data race is not won by chance but by design.

What the Companies Winning at Big Data Do Differently

Few industries illustrate the Big Data wars better than the media business. In TV programming, combatants like Netflix and Amazon challenge decades-old premium channels and other producers to gain watchers with original programming. Using their treasure troves of information on online customer viewing habits, they’re designing new TV series that their data tells them will win.
Early results show that it is working — and that many pre-Web media companies should be concerned. Netflix’s first foray into original programming, the “House of Cards” series, has been a big hit (although the company doesn’t disclose ratings). The company says the series brought in 2 million new U.S. subscribers in the first quarter of 2013, a 7% increase over the previous quarter. The company’s chief content officer said in February that Netflix uses “really big data” to pick which shows to produce and how to promote them. Yet Netflix’s debut as a producer also demonstrates something far more important: how Big Data can fundamentally change the structure of an industry by shifting the balance of power.
The Big Data wars are hardly limited to the media industry. In December 2012 and January of this year, Tata Consultancy Services surveyed 1,217 executives from large companies (revenue of more than $1 billion) in a dozen global industries in North America, Europe, Asia-Pacific, and Latin America. We found that companies with huge investments in Big Data are generating excess returns and gaining competitive advantages, putting companies without significant investments in Big Data at risk. The reason: There’s a big learning curve with Big Data, one that companies such as Netflix and Amazon had to embrace in the 1990s to deal with hundreds of millions of customer clicks.
It’s a learning curve that most other large companies have not yet faced. So what differentiated the companies with the greatest expected returns on Big Data for 2012 from those with the smallest? We categorized as “leaders” survey respondents that estimated a greater than 50% return on their Big Data investments last year (a number far above most companies’ hurdle rate). Although they tended to make much larger investments in Big Data, they also generated much higher returns on those investments than the laggards did. Higher spending correlates with more headroom for revenue growth.
Specifically, our study found that the companies estimating the greatest returns last year on Big Data outspent those with much smaller ROI by a factor of more than three — a median spend of $24 million vs. $7 million. Of the 53% of the survey respondents that had Big Data initiatives in 2012, median spending per company was $10 million, a relatively small amount given that median revenue was $6.9 billion. But that median spending masks a great polarity in Big Data investments — a huge gulf between companies that have embraced it and those that are slow to adopt. A narrow slice of the respondents (7%) with Big Data initiatives in 2012 invested at least $500 million each on Big Data software, hardware, data scientists, consultants and other related expenses. On the other end of the spectrum, 24% spent less than $2.5 million apiece on it last year.
Of course, simply spending more is not a strategy. How else did the companies projecting the greatest ROI from Big Data differ from under-performers? We found four key differences:

  • Leaders are more Internet-centric. On average, 42% of total revenue of the ROI leaders came from customer orders received via the Internet, compared to just 29% for the laggards. This may not be a big surprise; many of the early Big Data technologies such as the Hadoop Big Data system came from Internet companies themselves such as Yahoo and Google. Internet companies face many digital interactions, so they need Big Data technologies and people to sort through their click-stream data. Yet ROI leaders in our study also included telecoms, retailers, banks and high-tech companies. You don’t have to be an Internet company to generate outsized returns on Big Data.
  • Leaders panned for gold in several places. ROI leaders see greater potential from Big Data to improve a number of marketing, sales, R&D and service activities. Companies such as Procter & Gamble and Netflix are using Big Data to identify new product opportunities. Leaders also believe Big Data holds much greater potential than do under-performers for improving four marketing activities: monitoring and improving customers’ experience in offline channels (such as stores); discerning competitors’ moves beyond pricing; monitoring external perceptions of the brand; and marketing based on customers’ physical location (which is why it’s become important for many companies to buy mobile data). This last activity helps explain the appeal of Big Data to a growing number of retailers. Leaders also see big potential from Big Data to monitor their products’ performance in the field. General Electric is committing $1 billion to an analytics and software center to monitor the performance of its aircraft engines, healthcare equipment, power generation and other machinery, helping customers get more value.
  • Leaders are more aggressive in exploiting unstructured and external data. Unstructured data, or digitized text, video, machine and other data that doesn’t easily fit into traditional databases, is hard for computers to analyze. That’s changing, however, as analytics tools come to market for performing such compute-intensive chores as discerning sentiments from text. The ROI leaders make such unstructured data a bigger part of their data mix (55% of their digital data is unstructured or semistructured) vs. 46% for the ROI laggards. And on another measure, leaders use a higher percentage of external data (that is, data they don’t own): 37% vs. 26% for laggards. Retailers that want mobile location data of their shoppers must get that “external” data from the telcos.
  • Leaders are more likely to create a home for their Big Data professionals. Instead of embedding data scientists in business functions, ROI leaders centralized their analysts. Some 79% of the ROI leaders put their analysts in a dedicated Big Data group or in IT vs. 68% of laggards. The manager of a large analytics team at a big Internet company believes that removing analysts from business units and functions and centralizing them was critical to success. When they reported to business unit managers, “our analysts got heavy pressure to confirm what those unit managers were already doing,” the manager said. Centralizing the analysts also helped them share analytics methods, which he termed the “special sauce.” Providing unvarnished advice to key executives about how to optimize the website, the analysts have helped the firm increase revenue by tens of millions of dollars.

One cautionary note: while Big Data can help businesses identify unnecessary costs, cost-reduction strategies face diminishing returns. Companies such as Netflix, General Electric and LinkedIn have revealed far more potential in using Big Data in sales, marketing, R&D and other revenue-generating activities to drive growth.
Companies still dabbling in “small data” would be well-advised to make serious changes or risk losing to those that have adopted Big Data in a big way.